Radiative skin temperature is often used to examine heat exposure in multi-city studies and for informing urban heat management efforts since urban air temperature is rarely measured at the appropriate scales. Cities also have lower relative humidity, which is not traditionally accounted for in large-scale observational urban heat risk assessments. Here using crowdsourced measurements from over 40,000 weather stations in ≈600 urban clusters in Europe, we show the moderating effect of this urbanization-induced humidity reduction on heat stress during the 2019 heatwave. We demonstrate that daytime differences in heat index between urban clusters and their surroundings are weak and associations of this urban-rural difference with background climate, generally examined from the skin temperature perspective, is diminished due to moisture feedback. We also examine the spatial variability of skin temperature, air temperature, and heat indices within these clusters, relevant for detecting hotspots and potential disparities in heat exposure, and find that skin temperature is a poor proxy for the intra-urban distribution of heat stress. Finally, urban vegetation shows much weaker (~1/6^th as strong) associations with heat stress than with skin temperature, which has broad implications for optimizing urban heat mitigation strategies. Our results are valid for both operational metrics of heat stress (such as apparent temperature and Humidex) and for various empirical heat indices from epidemiological studies. This study provide large-scale empirical evidence that skin temperature, used due to the lack of better alternatives, is weakly suitable for informing heat mitigation strategies within and across cities, necessitating more urban meteorological observations.